Rosetta is molecular modeling software that has been developed for the prediction and design of macromolecular structure. Rosetta has performed well in community wide assessments of protein structure prediction and docking, and it has been used to design new protein structures as well as design altered specificity protein-protein and protein-DNA interactions. Rosetta is developed and maintained by research groups at 11 separate universities, and over 2400 laboratories have obtained free licenses for the software. Until 2004, Rosetta was written in Fortran 77. Over the past three years we have used NIH support to create a new C++ object-oriented version of Rosetta. The new software is organized as a set of libraries that contain classes and routines for representing and scoring molecular systems, for holding move sets, and for performing optimization of macromolecular conformation and sequence. The primary goal of this proposal is to capitalize on this rewrite, and extend the Rosetta software in directions that were less feasible when the code was not modular and object-oriented. Currently, Rosetta runs through a command line interface that forces users to choose from a fixed set of modeling protocols, such as protein-protein docking or the design of protein monomers. We will create a framework within Rosetta that allows users to easily create custom protocols in either C++ or with the scripting language Python. The python binding will be used to create an interface with PyMOL, a widely used program for molecular visualization. This will provide a graphical interface for initiating and interacting with Rosetta simulations as well as rapidly evaluating the quality of Rosetta models. Most applications of Rosetta benefit significantly from increased sampling of conformational and/or sequence space, and therefore, benefit from faster algorithms. We will increase the speed of Rosetta calculations by taking advantage of new C++ objects for caching energies. This grant will also support developers and users by maintaining benchmarks that test the integrity of the code, maintaining the user's guide and supporting meetings between developers at the various institutions. PUBLIC HEALTH RELEVANCE: The function of a protein, RNA or DNA molecule is largely determined by its 3- dimensional structure. We aim to develop a state-of-the-art computer program for predicting and designing the structures of biological macromolecules. The program will be freely available to academic laboratories and its predictions will help investigators understand and fight human diseases such as cancer and AIDS.